Investigating the numerical parameter space for a stenosed patient‐specific internal carotid artery model

Abstract

Systemic risk factors are known to correlate with cardiovascular diseases, but e.g., atherosclerotic plaques are focally distributed and highlight the role of hemodynamically induced forces on vascular remodeling. Computational fluid dynamics (CFD) shows great promise for revealing mechanisms of atherosclerotic plaque progression, but the utility of CFD depends on the robustness of the numerical methods. The aim of the study was to investigate the parameter space of the numerical solutions to understand the resulting flow effects in a stenosed patient-specific internal carotid artery model. Simulations were performed on meshes consisting of 2 to 50M-elements meshes with a kinetic energypreserving and minimally-dissipative solver, and time step size ranging from 1 ⋅ 10$% to 5 ⋅ 10$' seconds. The spatial refinement study revealed large differences in the instantaneous velocity fields, and the coarsest simulation did not provide any meaningful insight into the flow. That being said, the time-averaged results were in acceptable agreement for all spatial and temporal refinement levels. The variations in temporal resolution had minor effects, and the coarsest resolution was found to suffice. In conclusion, even for a highly accurate solver, a relatively high spatial resolution was needed to sufficiently resolve the flow, and we found the 22M-element mesh to offer an optimal balance between computational cost and time-averaged quantities.